Objective: The presence of atypical cells in urine cytology is unsatisfactory for both cytologists and clinicians. The objective of this study was to test whether p53 and Ki-67 immunostaining could improve urothelial carcinoma (UC) detection on urinary cytology.
Methods: A total of 196 urine samples were analysed, 142 from the bladder, 41 from the upper tract and 13 from ileal bladder replacement. Cytology results were expressed as normal (N) (n = 81), atypia cannot exclude low-grade UC (ALG) (n = 25), suspicious for high-grade UC (SHG) (n = 39) and high-grade UC (HG) (n = 51). Actual diagnoses were confirmed by histopathological analysis, cystoscopic examination or follow-up for at least 1 year. Immunocytochemistry performed on CytoSpin™ slides allowed the determination of the percentage of positive cells with p53 and Ki-67.
Results: The median percentage values [first to third quartile] of p53 and Ki-67 were 0 [0-5] and 0 [0-1] for N cytology, 5 [0-40] and 2 [1-10] for ALG, 10 [0-30] and 6 [3-25] for SHG, and 30 [10-80] and 20 [10-30] for HG, respectively. Statistically higher values were observed for both tests (P < 0.001) in positive cytologies (ALG, SHG and HG). The optimal cut-offs were 5% for p53 and 3% for Ki-67. The sensitivity and specificity for the detection of all UC were 86.4% and 76.7% for cytology alone, 81.3% and 93.2% for cytology and p53, 75.7% and 88% for cytology and Ki-67, and 68.9% and 97.5% for cytology, p53 and Ki-67, respectively.
Conclusion: Using p53 and/or Ki-67 in addition to cytology increases the specificity without penalising the sensitivity.
Keywords: Ki-67; immunocytochemistry; p53; urine cytology; urothelial carcinoma.
© 2016 John Wiley & Sons Ltd.